Many children with specific language impairment (SLI) show serious limitations in grammatical ability. These limitations are often longstanding, and they represent one of the most reliable indicators of this disorder. An especially common manifestation of this grammatical weakness is the inconsistent use of grammatical morphemes that mark tense and agreement (e.g., -s in jumps and is in The girl is eating). Although the details of this inconsistent use have been well described in the scientific literature, an explanation for the protracted period of inconsistency has not yet emerged. The goal of this project is to test a new proposal regarding the possible source of the inconsistent use of tense and agreement morphemes by children with SLI. Experiments 1-5 are designed to test the hypothesis that children with SLI have difficulty interpreting sentences that contain subject-verb sequences lacking tense and agreement morphology (as in the girl go in the question Did the girl go? or the boy running in the sentence I see the boy running). As a result, they treat these sequences (e.g., The girl go) as grammatical and use them in their own speech. [Furthermore, because the sentence interpretation problems are still present when the children begin to form abstract rules for sentence use, generalizations from these improperly extracted sequences become the source for generating new utterances, thus prolonging the period of inconsistent use of tense and agreement markers.] The first five experiments are designed: (a) to identify the specific types of sequences that might be problematic (e.g., questions and/or complex sentences); and (b) to determine whether the children generalize misinterpreted sequences to form new (ungrammatical) utterances of the same type (e.g., does misinterpretation of the girl go in Did the girl go? lead to spontaneous generation of productions such as The cat go?). [In addition, these experiments include complex sentence comprehension and working memory tasks as predictors to determine the relative contributions that syntactic comprehension difficulty and working memory limitations make to the children's tendency to omit tense and agreement morphology.] Experiment 6 is an intervention study designed to determine whether the period of inconsistency with tense and agreement morphology can be significantly reduced by manipulating the language input the children receive. Specifically, it is predicted that inconsistency can be reduced by exposing the children to many sentences containing the target morphemes (e.g., The girl jumps, The boy is walking) and minimizing exposures of questions (e.g., Did the girl go?) and complex sentences (e.g., I see the boy running) that contain subject-verb sequences with no tense and agreement morphology. [Along with its clinical value, this experiment serves as a critical test of the causal role of input distribution on the children's sentence use.] The findings of this project should contribute to a greater understanding of the sources of grammatical limitations in children with SLI and provide important information concerning how best to treat the pervasive tense and agreement weaknesses seen in this disorder.